Modifying Akt Signaling in B-Cell Chronic Lymphocytic Leukemia Cells

Published OnlineFirst September 7, 2010; DOI: 10.1158/0008-5472.CAN-09-4411 Published OnlineFirst on September 7, 2010 as 10.1158/0008-5472.CAN-09-4411

Tumor and Stem Cell Biology Cancer Research Modifying Akt Signaling in B-Cell Chronic Lymphocytic Leukemia Cells

Sebastian W. Hofbauer1, Josefina D. Piñón1, Gabriele Brachtl1, Lucia Haginger1, Wei Wang1, Karin Jöhrer2, Ingeborg Tinhofer1,3, Tanja Nicole Hartmann1, and Richard Greil1

Abstract Emerging evidence suggests that the survival of B-cell chronic lymphocytic leukemia (CLL) cells is dependent on microenvironmental influences such as antigenic stimulation and support by stromal cells. Akt, also known as protein kinase B, is a central component in prosurvival signaling downstream of these events. We investigated the role of Akt and its modulation by the protooncogene T-cell leukemia 1a (Tcl1a) in the survival pathways of primary CLL samples and CLL-derived prolymphocytic cell lines MEC-1 and MEC-2. Akt activation was increased by the protective presence of human bone marrow stromal cells and B-cell receptor mimicking signals but antagonized by direct Akt blockade with the novel specific inhibitor AiX, with preferential apoptosis induction in CLL cells with an unmutated immunoglobulin status, which predicts poor clinical outcome. In addition, we found a direct interaction of Akt with Tcl1a in an endogenous coimmunoprecipitation assay. Confirming the critical role of Tcl1a in modulating Akt signaling, Akt activation was enhanced by overexpressing Tcl1a in CLL. In contrast, decreasing Tcl1a levels by small interfering RNA reduced Akt activation in the fludarabine-insensitive CLL cell line MEC-2 and sensitized the malignant cells to fludarabine treatment. In summary, our data reveal a significant role for the Akt-Tcl1a axis in CLL survival and propose a further evaluation of this interplay for targeting chemoresistance phenomena. Cancer Res; 70(18); 7336–44. ©2010 AACR.

Introduction erally follow a more indolent course than those with unmutated (UMT) IgVH genes, who also tend to show evidence of B-cell chronic lymphocytic leukemia (CLL) is the most adverse cytogenetic features, clonal evolution, and resistance prevalent non–Hodgkin lymphoma in the Western world to therapy (2, 3). In addition, molecular markers that are and is characterized by the accumulation of CD5+ B lympho- associated with the signaling capacity of the BCR in CLL, cytes in the blood, lymph nodes, and bone marrow (BM). The such as CD38 and the ζ-associated protein 70 (ZAP-70), serve clinical course and outcome of CLL varies considerably. as indicators for poor clinical prognosis (reviewed in ref. 4). Some patients progress rapidly and die early despite therapy, In normal B lymphocytes, BCR stimulation triggers the whereas others exhibit an indolent disease course and a nor- recruitment and activation of Syk kinases, followed by the mal life span (1). Emerging evidence suggests that the B-cell activation of effector enzymes, including protein kinase C antigen receptor (BCR) is the major determinant of this dif- (PKC), phosphatidylinositol 3-kinase (PI3K), and phospholi- ferential outcome. In particular, patients carrying mutated pase Cγ2. These signaling complexes regulate key down- (MUT) variable heavy chain region genes encoding the stream pathways such as extracellular signal-regulated surface immunoglobulin component of the BCR (IgVH) gen- kinase, c-Jun NH2-terminal kinase, p38, and Akt, the out- comes of which determine the B-cell fate (5). The serine/ threoninekinaseAkt,alsoknownasproteinkinaseB,isa Authors' Affiliations: 1Laboratory for Immunological and Molecular Cancer Research, University Clinics of Internal Medicine III with central molecule in the transmission of antigenic BCR- Hematology, Oncology, Hemostaseology, Infectiology and derived signals. Several studies suggested that Akt activation Rheumatology, Salzburg, Austria; 2Tyrolean Cancer Research Institute, Innsbruck, Austria; and 3Translational Radiobiology and Radiooncology promotes CLL cell survival following BCR engagement (6, 7). Research Laboratory, Clinical Department for Radiotherapy (CCM/CVK), It is, however, controversially discussed whether in peri- Charité University Hospital, Berlin, Germany pheral blood CLL cells, constitutively active Akt contributes Note: Supplementary data for this article are available at Cancer Research to cell survival (6, 8–11). Online (http://cancerres.aacrjournals.org/). The protooncogene T-cell leukemia 1a (Tcl1a) is an Corresponding Author: Richard Greil, Laboratory for Immunological and important modulator of Akt activity. Tcl1a was initially char- Molecular Cancer Research, University Clinics of Internal Medicine III with Hematology, Oncology, Hemostaseology, Infectiology and Rheumatolo- acterized in preleukemic T cells (12) but is also expressed in gy, Salzburg, Austria. Phone: 43-662-4482-2880; Fax: 43-662-4482- subsets of naïve lymphocytes (13). Tcl1a protein is expressed 2898; E-mail: [email protected]. in 90% of human CLL samples at variable levels correlating doi: 10.1158/0008-5472.CAN-09-4411 with ZAP-70 expression, IgVH mutational status, and cytoge- ©2010 American Association for Cancer Research. netic aberrations (14). Tcl1a can form a hetero-oligomeric

7336 Cancer Res; 70(18) September 15, 2010

CLL Apoptosis by Akt Modulation

protein complex with Akt, thereby serving as a coactivator of Analysis of cell viability, Tcl1a expression, and Akt (15, 16). The significance of the Akt-Tcl1a axis is further Akt activation underlined by the fact that TCL1A-transgenic mice expand a Cells were stained with fluorochrome-labeled anti-CD5, CD5+/IgM+ lymphocyte population, which is reminiscent of anti-CD19, Annexin V, and 7-aminoactinomycin D (7-AAD), human CLL (17) and is accompanied by increased Akt acti- and at least 10,000 CLL cells were analyzed using a FC-500 vation levels (18). Intriguingly, these mice have a high-risk flow cytometer and CXP2-2 software (Beckman-Coulter). phenotype resembling the typical molecular features of an For intracellular phospho-Akt (Ser473)andTcl1astaining, UMT BCR of human CLL cells. cells were fixed and permeabilized using the Fix and Perm Direct interference in Akt activation by specific Akt inhibi- kit (AN DER GRUBB) and stained with isotype control, tors or modulation of the coactivator Tcl1a has not been anti-Tcl1a, anti–phosphoserine Akt, or anti–pan-Akt (19). extensively investigated thus far. We analyzed these aspects in the context of BCR-related signaling and the tumor Plasmid construction and transfection microenvironment—two elements centrally involved in surviv- TCL1A wild-type cDNA was PCR amplified and subcloned al and progression of CLL. We found a significant role for the into pcDNA3.1 vectors (Clontech). For cloning, the following Akt-Tcl1a axis in CLL cell survival. Our results warrant a fur- PCR primers with EcoRI and SalI restriction sites were used: ther evaluation of this interplay as a therapeutic target in CLL. GTGGAATTCGCCATGGCCGAGTGCCCGACACTCGGG (forward) and GTGGTCGACTCAGTCATCTGGCAGCAGCTC- Materials and Methods GAGAAG (reverse). Plasmids were nucleofected into MEC-1 cells using the Human B Cell Nucleofector kit (Lonza Reagents and antibodies Cologne AG). Transfected cells were selected by geneticin. The Akt inhibitor 10-(4′-(N-diethylamino)butyl)-2-chloro- phenoxazine, HCl (AiX) was purchased from Calbiochem, Small interfering RNA transfection the PI3K inhibitors LY294002 and U0126 were from Cell Sig- MEC-2 cells were transfected with 100 pmol small interfer- naling Technology, and 2-fluoroadenine-9-β-D-arabinofura- ing RNA (siRNA) using Metafectene Pro (Biontex GmbH) noside and rottlerin were from Sigma-Aldrich. Antibodies according to the manufacturer's instructions. Assessment against Akt1, pan-Akt, phospho-Akt (Ser473), phosphoserine of transfection efficacy was monitored by transfecting an glycogen synthase kinase 3α/β (GSK3α/β), Mcl-1, Tcl1a, Alexa Fluor 647–conjugated nontargeting siRNA and flow Bid, and Bik were purchased from Cell Signaling Technology; cytometric analysis, and effective protein knockdown was antibodies against Akt2, MDM2, p53, and phospho–Bcl-2 densitometrically assessed. (Ser70) were from Santa Cruz Biotechnology; Puma and tubulin were from Sigma-Aldrich; antibodies against CD5-PC7, Immunoblot analysis CD19-phycoerythrin (PE), and CD19-PC7 were from Beckman- Standard lysis buffer was supplemented with phosphatase Coulter; VLA-4 (CD49d)-PE was from BD Biosciences; and (Roche) and protease inhibitors (Sigma-Aldrich). For coim- CD38-PE was from Immunotech. munoprecipitation, a buffer containing 0.3% CHAPS was used (20). The lysate protein concentration was determined Patients, cell preparation, and AiX treatment using the Bio-Rad protein assay. For coimmunoprecipitation, Following informed consent, peripheral blood samples lysates were precleared with 50 μL protein A or G dynabeads were obtained from patients fulfilling diagnostic and immu- (Invitrogen) and further incubated with the antibody over- nophenotypic criteria for CLL at the Salzburg University Hos- night. Magnetically captured complexes were washed and pital, Austria (Supplementary Table S1). Peripheral blood analyzed by immunoblotting. Signals were detected by mononuclear cells (PBMC) were isolated by density gradient chemiluminescence using the enhanced chemiluminescence centrifugation and cultured in complete RPMI 1640. To gen- kit (Millipore). erate primary BM-derived stromal cells, adherent mononuclear cells from CLL BM aspirates were cultured for 3 weeks in Statistical analysis RPMI 1640 supplemented with 1 μmol/L hydrocortisone and Statistical analysis was performed using GraphPad Prism 100 μmol/L β-mercaptoethanol. For coculture experiments 5.0 or SPSS 17.0. All data were tested for normal distribution, including risk stratification, BM-derived stromal cells from and t test or ANOVA and Bonferroni post test were healthy donors were kindly provided by Reinhard Henschler performed for normally distributed data. For nonnormally (Institute for Transfusion Medicine and Immune Hematology, distributed data, the Mann-Whitney test was used. German Red Cross Blood Donor Service, Frankfurt, Germany). Results Cell lines HS-5 stromal cells and C7H2-CEM cells were purchased Direct Akt inhibition induces CLL apoptosis that from the American Type Culture Collection, and MEC-1 cannot be compensated by protective stimuli and MEC-2 cells were from the German Resource Centre We first tested the in vitro effect of the novel Akt inhibitor for Biological Material. Cell lines were cultured according AiX on CLL survival. AiX has been previously shown to effec- to the supplier's instructions, and cell aliquots from early tively inhibit Akt and Akt downstream target phosphorylation passages (P1–P10) were used. without influencing the activity of PI3K and other Akt-related

www.aacrjournals.org Cancer Res; 70(18) September 15, 2010 7337

Hofbauer et al.

kinases (21). We treated randomly selected CLL samples mal DNA fragmentation, and caspase-3 activity as commonly with increasing concentrations of AiX for 24 hours and used indicators for apoptosis. Akt inhibition resulted in high subsequently assessed CLL cell viability by flow cytometry fol- levels of cleaved PARP and DNA ladder formation, suggesting lowing Annexin V/7-AAD staining. AiX decreased Akt phos- induction of apoptosis rather than necrosis (Supplementary phorylation and CLL cell viability in a dose-dependent and Fig. S2A and B). We also observed increased caspase-3 activity effective manner (Fig. 1A; data not shown) with an IC50 value in CLL cells after AiX treatment, which could be mostly of 4.4 μmol/L. Comparison of apoptosis rates induced by blocked by preincubation with the pan-caspase inhibitor either AiX or the PI3K inhibitor LY294002 showed similar Q-VD-OPH (Supplementary Fig. S2C). effectivity at 5 μmol/L but dramatically higher effectivity of CLL cell survival ex vivo critically depends on the presence AiX at 10 μmol/L (P < 0.001; Supplementary Fig. S1). For dis- of supportive accessory cells such as BM-derived stromal crimination between AiX-induced apoptosis and necrosis, we cells. Conventional in vitro culture conditions without these additionally monitored for caspase-dependent cleavage of the accessory cells usually result in a rapid decline in CLL cell poly(ADP-ribose) polymerase (PARP) protein, internucleoso- viability. Concordantly, we observed that CLL cells cocultured

Figure 1. AiX-induced CLL apoptosis cannot be antagonized by the presence of protective stromal cells. A, PBMCs from CLL patients were cultured in the presence of increasing concentrations of AiX for 24 h and cell viability was cytometrically measured after Annexin V/7-AAD staining. CLL cells were hereby defined by gating on CD19+/CD5+ expression. The median with min/max of eight CLL samples is shown, and statistically significant differences are indicated. **, P < 0.01; ***, P < 0.001. B, left, PBMCs from CLL patients were cultured in the presence of HS-5 stromal cells for 5 d. CLL cell viability was defined as described above (n = 5). Right, Akt activation in CLL cells gained by the presence of HS-5 cells was cytometrically measured as the ratio of phospho-Akt (Ser473)/Akt in an intracellular stain. C, left, PBMCs from eight patients were cultured in the presence or absence of CLL BM-derived stromal cells and treated with 5 μmol/L AiX for 72 h. Right, stromal cells were treated with 5 μmol/L AiX for 72 h. CLL and stroma cell viability was measured as described (median with min/max). *, P < 0.05; **, P < 0.01. n.s., not significant.

7338 Cancer Res; 70(18) September 15, 2010 Cancer Research

CLL Apoptosis by Akt Modulation

Figure 2. PKCβ-dependent Akt activation enhances CLL cell viability and fludarabine resistance. A, cells were stimulated with 10 nmol/L PMA for 20 min in the presence or absence of the indicated inhibitors. The activation status of Akt was monitored by immunoblotting the lysates for phosphoserine Akt (pS-Akt) and total (pan-) Akt. Note the very short exposure time. B, PBMCs from CLL patients were cultured for 48 h with 10 nmol/L PMA, 10 μmol/L AiX, or both (n = 7). Cell viability was measured as described (median with min/max). *, P < 0.05. C, CLL cells were treated with 10 nmol/L PMA and 10 μmol/L AiX for 6 h. Cell lysates were separated by SDS-PAGE, and expression status of Mcl-1, phosphoserine GSK3α/β, Tcl1a, MDM2, p53, Puma, phospho–Bcl- 2 (Ser70), total Bcl-2, Bid, and Bik was monitored by immunoblotting. Tubulin expression is shown as equal protein loading control. D, PBMCs from CLL patients were activated with 10 nmol/L PMA and treated with 5 μmol/L fludarabine. After 48 h, cell viability was assessed as described. The median with min/max of five independent experiments is shown. *, P < 0.05.

with the human stromal cell line HS-5 displayed significant Besides the stromal microenvironment, antigenic input is increased survival (Fig. 1B, left). Notably, this stromal- one of the most relevant factors in CLL survival (22). We induced prosurvival effect correlated with increased Akt recently showed a critical role of PKCβ, a molecule centrally phosphorylation in the same CLL cells (n = 5, correlation involved in BCR signaling, in CLL development (23). We now coefficient r2 =0.81,P = 0.032; Fig. 1B, right) and could be stimulated CLL cells with phorbol 12-myristate 13-acetate completely countered by treatment with AiX (data not (PMA), thereby mimicking a PKCβ signal (24). This stimula- shown). As we wished to determine if Akt inhibition of the tion resulted in strong Akt phosphorylation at Ser473, a major CLL cells or rather the reduced viability of the stromal cells and for full activation indispensable phosphorylation site (25), accounted for the decline in stromal-mediated protective which was entirely reversible by AiX treatment (Fig. 2A). We effects, we assessed the viability of the HS-5 cells on AiX confirmed the PKCβ mimicking nature of PMA stimulation treatment. We observed a slight reduction in HS-5 viability by use of enzastaurin. This PKCβ-specific inhibitor, but not in response to ≥5 μmol/L AiX (data not shown). Therefore, the PI3K inhibitor LY294002 or the PKCθ inhibitor rottlerin, we repeated these experiments using primary patient- antagonized PMA-induced Akt activation (Fig. 2A), as previ- derived BM stromal cells, which were resistant to AiX treat- ously observed (23). PMA stimulation led to a significant ment. Strikingly, coculture of CLL cells with primary stromal increase in CLL cell viability but failed to inhibit AiX-induced cells almost completely protected CLL cells against sponta- apoptosis (Fig. 2B). Determining the molecular basis of this neous apoptosis during the whole culture period of 5 days observation, we found a dramatic activation of the Akt (Fig. 1C, left). The median viability of CLL cells after 5 days signaling pathway on PMA stimulation, such as increased of in vitro culture in the absence of stromal cells was 25.3% levels of phosphoserine GSK3α/β and strong upregulation compared with 91.4% in the presence of the CLL BM- of the antiapoptotic protein Mcl-1, which was mostly antag- derived stromal cells. However, despite this strong pro- onized by AiX treatment. Expression of the direct Akt tective effect, Akt inhibition was highly efficient in inducing modulator Tcl1a was not visibly affected by PMA stimulation CLL apoptosis without affecting stromal cell viability or direct Akt inhibition (Fig. 2C). Importantly, Akt inhibition (Fig. 1C, right). was efficient in antagonizing PMA-induced Bcl-2 Ser70

www.aacrjournals.org Cancer Res; 70(18) September 15, 2010 7339

Hofbauer et al.

phosphorylation, recently reported to be crucial for CLL sur- we observed clear apoptosis induction by AiX. All samples vival (26). In addition, proapoptotic BH3-only proteins Bid displayed a 17p13 deletion in 40% to 90% of the cells, and Bik were upregulated in the presence of Akt inhibition resulting in p53 dysfunction. Importantly, AiX treatment (Fig. 2C). was similarly effective in these fludarabine-resistant cells We additionally checked the involvement of the p53 sig- (n = 4) and normal samples (n =7,P = 0.9972; Supplemen- naling pathway because p53 dysfunction is a primary tary Fig. S3). determinant of CLL chemoresistance (27, 28). Furthermore, protein expression of MDM2, the main counter player of High-risk CLL cells are more sensitive toward p53, is regulated by Akt signaling (29, 30). MDM2 expres- Akt inhibition than low-risk cells sion was increased on PMA stimulation but decreased with We further analyzed AiX-induced apoptosis in different additional AiX treatment, coinciding with apoptosis. CLL subgroups. Stratification of patients' samples according Expression levels of p53 and its downstream target Puma to their molecular risk profile (marked in Supplementary were also increased on PMA. The observation that we Table S1) revealed that in solo culture, CLL cells with an could modulate p53 pathway members by PMA or AiX UMT IgVH status were significantly more sensitive to treat- treatment led us to investigate the influence of the ment with AiX than those with a MUT status (Fig. 3A, left). PKCβ/Akt axis on the fludarabine responsiveness of CLL Furthermore, AiX also preferentially killed CLL cells with cells. Whereas fludarabine treatment resulted in a signifi- high CD38 and VLA-4 expression (Fig. 3B and C, left), consis- cant decrease in cell viability of unstimulated cells, the tent with the association of these molecules with poor clin- same concentration did not affect the viability of PMA- ical outcome (2, 31, 32). Spontaneous apoptosis during stimulated cells, further supporting the involvement of in vitro culture did not differ between CLL subgroups. In the PKCβ/Akt/Mcl-1 axis in CLL survival and chemoresis- the presence of stromal cells, Akt inhibition was very effec- tance (Fig. 2D). Concordantly, when we tested the sensitiv- tive with a trend of preferential targeting UMT but not VLA- ity of fludarabine-resistant CLL samples to Akt inhibition, 4–high cases (Fig. 3C, right). Due to this observation, we

Figure 3. CLL cells with UMT IgVH genes and high CD38 or VLA-4 expression are highly sensitive to Akt inhibition. Cells were treated with AiX either in solo culture (5 μmol/L AiX, 24 h; left) or in the presence of stromal cells (10 μmol/L AiX, 48 h; right), and CLL cell viability was determined as described. Patient samples were stratified according to the mutational status of the IgVH genes (A), CD38 expression (B), or VLA-4 expression (C). Significance levels and analyzed case numbers are indicated in the diagrams. *, P < 0.05, one-way ANOVA. D, Akt substrate phosphorylation was monitored in protein lysates from eight MUT and nine UMT cases. Mean densities of Akt substrate phosphorylation were normalized to the housekeeping protein tubulin. Tcl1a expression was analyzed by intracellular staining and fluorescence-activated cell sorting analysis in 16 UMT CLL and 14 MUT CLL cases.

7340 Cancer Res; 70(18) September 15, 2010 Cancer Research

CLL Apoptosis by Akt Modulation

determined basal Akt substrate phosphorylation and Tcl1a levels in UMT versus MUT CLL subgroups. As expected, UMT CLL cells significantly displayed higher Akt activity and Tcl1a levels than MUT CLL cells (Fig. 3D).

The Tcl1a/Akt interaction axis is involved in fludarabine resistance of a CLL-derived prolymphocytic cell line The expression of Tcl1a reciprocally correlates with the IgVH mutational status and is centrally involved in CLL tumorigenesis (14, 33, 34). Furthermore, Tcl1a is an important modulator of Akt activity with a central role in transmission of antigenic BCR-derived signals (35). Having observed significantly higher Tcl1a levels in UMT than MUT CLL cells, we next addressed the role of the Tcl1a/Akt interplay in CLL survival and chemosensitivity. Therefore, we used MEC-1 and MEC-2 CLL-derived prolymphocytic cells displaying different Tcl1a expression levels with higher Tcl1a expression in MEC-2 cells. Increased Tcl1a expression correlated with increased levels of phosphoserine Akt and GSK3 in MEC-2 cells (Fig. 4A). We tested the sensitivity of these cells toward Akt inhibition. AiX decreased MEC cell viability in a dose-dependent manner (Fig. 4B). After treatment with 20 μmol/L AiX, 50% of the MEC-1 cells and 65% of the MEC-2 cells underwent apoptosis (a fraction of 0.5 and 0.65 compared with untreated control cells, respectively). To determine whether this higher constitutive Akt phosphorylation of MEC-2 is a consequence of Tcl1a/Akt interaction, we next examined for direct Tcl1a/Akt protein-protein interaction. Three isoforms of Akt have been described: Akt1, Akt2, and Akt3. We observed Akt1 and Akt2 expression in MEC-1 and MEC-2 cells but did not detect Akt3 (Fig. 4C; data not shown). Performing endogenous coimmunoprecipitation experiments, we observed direct protein-protein interaction of Tcl1a and Akt (Fig. 4C). Tcl1a coimmunoprecipitated with both Akt1 and Akt2 isoforms in lysates of the Tcl1a-high CLL cell line MEC-2. No coimmunoprecipitation of Tcl1a was Figure 4. Enhanced Tcl1a expression corresponds to increased Akt observed when lysates of the Tcl1a-negative CEM cell line activation. A, MEC-1 and MEC-2 cell lysates were separated by – c SDS-PAGE and expression levels of Tcl1a, phosphoserine Akt, and or anti cytochrome IgG were used in control experiments. phosphoserine GSK3 were monitored by immunoblotting. One of three Although Akt1 is the more predominant isoform found in representative immunoblots is shown. B, MEC-1 and MEC-2 cells MEC-2 cells, quantitatively more Tcl1a was bound to Akt2 were treated with increasing concentrations of AiX (5–20 μmol/L) for 24 h, (Fig. 4C). Next, to investigate the direct effect of Tcl1a and cell viability was measured by an Annexin V/7-AAD staining and flow on Akt activity, we knocked down Tcl1a expression in cytometry. Points, mean of five independent experiments for each cell line; bars, SD. **, P < 0.01; ***, P < 0.001. C, CEM and MEC-2 cell MEC-2 cells by siRNA. The efficacy of the transfection proce- lysates were used to immunoprecipitate (IP) individual Akt isoforms under dure was cytometrically assessed by nontargeting siRNA. nondenaturing conditions. SDS-PAGE and immunoblotting were Twenty-four hours after transfection, cell viability was not performed for Tcl1a and pan-Akt as described in Materials and Methods. negatively affected and ∼90% of MEC-2 cells were trans- One of three representative immunoblots is shown. fected (Supplementary Fig. S4). This transfection rate trans- lated into a 70% knockdown of the Tcl1a protein (Tcl1a siRNA compared with nontargeting siRNA) and resulted in responses. In addition, given the high importance of decreased Akt and GSK3 phosphorylation (Fig. 5A). Follow- Akt-dependent signaling in CLL survival, we investigated ing the observation of impaired Akt signaling connected to a potential Tcl1a function in chemosensitivity of these reduced Tcl1a levels, we next investigated whether overex- cells. MEC-2 cells have a monoallelic deletion in 17p13 pressing Tcl1a would enhance Akt activation in MEC-1 cells. (36), resulting in fludarabine resistance based on attenua- A plasmid encoding the human TCL1A cDNA was introduced tion but not complete abrogation of the p53 pathway. We into MEC-1 cells. This resulted in an activation of Akt and confirmed residual p53 activity by showing increased its downstream target GSK3α/β (Fig. 5B), further confirm- expression of the p53 downstream target p21 following AiX ing the indispensable function of Tcl1a in Akt-mediated treatment (Supplementary Fig. S5). Finally, we performed a

www.aacrjournals.org Cancer Res; 70(18) September 15, 2010 7341

Hofbauer et al.

Tcl1a knockdown in the fludarabine-resistant MEC-2 cell line Discussion and then treated these cells with increasing concentrations of fludarabine. Tcl1a knockdown indeed sensitized the cells to- Akt is a central protagonist in signal transduction pathways ward the fludarabine treatment (Fig. 5C). involved in cell differentiation, migration, proliferation, and survival, with innumerable regulatory circuits operating within and between these functions (37). Moreover, Akt is involved in malignant cell transformation and in the development of tumor chemoresistance (38). For these reasons, modulation of Akt activity through inhibition of phosphatidylinositol-3 kinase (PI3K), its main upstream activator, has long been deemed an attractive therapeutic approach, and a range of different PI3K inhibitors has been developed in recent years for several tumor entities. However, thus far, lack of selectivity and solubility of these inhibitors has limited their use in clinical trials (37). Constitutive PI3K activation has been shown in freshly prepared CLL cells, suggesting that the PI3K/Akt axis is relevant in CLL pathogenesis (8, 9, 39). In addition, we and others showed that Akt can also be activated by PI3K- independent pathways, such as PKCβ (23, 39), recently shown to regulate the outcome of BCR signals and CLL tumorigenesis (23, 40). Likewise, stabilization of the Akt kinase activity by the oncoprotein Tcl1a (41) is believed to contribute to the development of the CD5+ B-cell malignancy observed in the TCL1A-transgenic mouse model (17, 18). Furthermore, this Tcl1a overexpression can partly compensate for the loss of PKCβ presumably by providing the missing Akt amplification signal in the malignant cell population (23). Surprisingly, although Akt-driven pathways play an important role in CLL tumorigenesis (6, 7, 10), the effects of direct Akt targeting or modulation of its interplay with its coactivator Tcl1a have not been widely addressed in CLL thus far. We show here that direct inhibition of Akt by AiX even under conditions that normally confer protection to these cells, such as survival signals emanating from the tumor microenvironment and signals mimicking sustained antigenic input, induced apoptosis of CLL cells. As CLL cells display an extended life span within the context of their immediate tumor microenvironment, but rapidly undergo spontaneous apoptosis ex vivo, increasing emphasis is being placed on these leukemic cell-microenvironment interactions (42). Concordantly, coculturing CLL cells with accessory cells (43) or activation of the BCR signaling pathway (7, 23) res- cues the malignant cells from succumbing rapidly to apoptosis ex vivo. Importantly, these prosurvival signals may converge on the Akt signaling pathway (6). Indeed, we show that both stromal- and BCR-related prosurvival stimuli result in Akt activation and that inhibition of Akt effectively over- comes these protective signals. The data indicate that direct Akt targeting can kill not only the bulk of the tumor cells circulating in the blood but also those residing within protec- Figure 5. Tcl1a expression is modulating Akt activation and fludarabine tive BM niches. Considering the pivotal role these niches play resistance of MEC cells. A, Tcl1a siRNA was performed in MEC-2 cells as described in Materials and Methods and the activation status of Akt was in the development of chemoresistance and disease relapse, monitored by immunoblotting. B, Tcl1a was overexpressed in MEC-1 cells as our results underscore the potential value of Akt inhibitors in described in Materials and Methods and the Akt activation status was the treatment of CLL. monitored after 48 h by immunoblotting. Immunoblots of A and B are Furthermore, we show that direct Akt inhibition leads to representative for three. C, MEC-2 cells were transfected with nontargeting (NT) or Tcl1a-specific siRNA and cultured in the presence or absence of strong proapoptotic effects especially in high-risk samples increasing concentrations of fludarabine for 48 h. Cell viability was measured and is also effective in chemoresistant cases. In the presence as described. Points, mean of three independent experiments; bars, SD. of primary stromal cells, Akt inhibition was very potent with

7342 Cancer Res; 70(18) September 15, 2010 Cancer Research

CLL Apoptosis by Akt Modulation

a trend of preferential targeting in those cases with an UMT therapeutic approach. Expression of Tcl1a has been observed IgVH status. In light of increased Tcl1a levels of these cells, in 90% of human CLL samples, with higher levels correlating we assume that their increased reliance on an activated Akt with an UMT IgVH status and ZAP-70 expression (14). More- pathway downstream of the BCR (6, 7) renders them more over, higher Tcl1a levels predict for an inferior clinical out- sensitive toward Akt inhibition. These risk-specific differ- come likely by modulating Akt activation (33). To show the ences in BCR signaling capacity could be overruled by PKCβ utility of a modulating approach, we made use of CLL cell– activation. Indeed, PKCβ stimulation with PMA resulted in derived lines that differed intrinsically in their expression high phospho-Akt levels as well as increased cell viability of Tcl1a. Concordantly, we found more Akt activation and and chemoresistance to fludarabine regardless of the risk higher AiX response rates with regard to cell death in high group. PMA-induced survival benefits were based on elevated Tcl1a-expressing MEC-2 cells than in MEC-1 cells (Fig. 4). levels of Mcl-1 and activated (phospho-)Bcl-2 (Fig. 2) with Of note, overexpression of myr-Akt1 in CLL cells increased simultaneous reduction of BimEL (data not shown). CLL leukemic cell size (6), which is an interesting observation, cells exist in a delicate balance between the proapoptotic as Akt-active MEC-2 cells have a bigger cell size than MEC- and prosurvival members of the Bcl-2 protein family. High 1 cells. Moreover, we are the first to report the coimmuno- levels of prosurvival Bcl-2 and Mcl-1, along with high levels precipitation of endogenous Tcl1a with Akt1 and Akt2 in the of proapoptotic Bcl-xL, Bax, Bim, and Bmf, have been CLL cell line MEC-2, as direct Tcl1a-Akt interaction has only observed in the majority of CLL cells (44, 45). Our data sup- ever been shown in cell lines engineered to overexpress Tcl1a port the described role of Mcl-1 and Bcl-2 in CLL cell survival (35). Condordant with our results, Tcl1a was shown to be and drug resistance (6, 7, 26). Furthermore, chemoresistance corecruited with pan-Akt in BCR-activated CLL cells (33). in CLL is largely dependent on loss of p53 function, in the RNAinterferencewithTcl1ainthesecellsledtoreduced form of 17p13 deletion, which also represents one of the Akt activation and enhanced response toward cytotoxic worst prognostic markers in CLL (46–48). Interestingly, we treatment such as fludarabine. In contrast, overexpression observed that activation of the BCR pathway using PMA of TCL1A in low Tcl1a-expressing MEC-1 cells led to led to increased protein levels of MDM2 and p53, as well enhanced phosphoserine Akt and GSK3α/β levels (Fig. 5). as induction of p53 target Puma. However, simultaneous Collectively, our data show the central role of the prosurvival activation of Akt and Mcl-1 by PMA could effectively sup- kinase Akt and its coactivator Tcl1a in the survival of stim- press the proapoptotic activity of these proteins (Fig. 2C). ulated CLL cells, which otherwise would be protected from In support of this, the strong downregulation of Mcl-1 by treatment with cytotoxic agents such as fludarabine, and Akt inhibition combined with simultaneous increase in proa- suggest that interference of this signaling axis represents a poptotic BH3-only proteins Bid and Bik resulted in apoptosis potential therapeutic strategy to combat CLL. even in PMA-stimulated CLL cells. Our observation on Bid cleavage on Akt inhibition may indicate the involvement of Disclosure of Potential Conflicts of Interest the extrinsic apoptotic pathway, as Bid conjoins the extrinsic and intrinsic apoptotic pathways via caspase-8. As we indeed No potential conflicts of interest were disclosed. observed caspase-8 activation after Akt inhibition (data not shown), it is likely that both apoptotic pathways are involved. Acknowledgments Our data are in line with coincidental Bid cleavage and caspase-8 activation after inhibition of the PI3K/Akt pathway We thank Mathias Lang for determination of caspase-3 activation. in different lymphoma entities (49, 50). Importantly, Akt inhibition also induced apoptosis in CLL Grant Support cells with p53 deletion or dysfunction (Supplementary Fig. S3), an observation in consensus with a recent report Klinische Malignom- und Zytokinforschung Salzburg-Innsbruck GmbH, SFB program P021 (R. Greil); ÖNB research grant 13420 (T.N. Hartmann); (11). Therefore, our results show that targeting Akt by specif- and grants of the province of Salzburg. ic small molecules represents a rational strategy to combat The costs of publication of this article were defrayed in part by the payment not only high-risk but also therapy-resistant CLL entities, of page charges. This article must therefore be hereby marked advertisement in including those with p53 dysfunction. accordance with 18 U.S.C. Section 1734 solely to indicate this fact. Besides direct Akt inhibition, disruption of the Akt inter- Received 12/03/2009; revised 07/06/2010; accepted 07/12/2010; published action with its coactivator Tcl1a can represent a valuable OnlineFirst 09/07/2010.

References 1. Hamblin T. Chronic lymphocytic leukaemia: one disease or two? Ann 4. Moreno C, Montserrat E. New prognostic markers in chronic lympho- Hematol 2002;81:299–303. cytic leukemia. Blood Rev 2008;22:211–9. 2. Damle RN, Wasil T, Fais F, et al. Ig V gene mutation status and CD38 5. Niiro H, Clark EA. Regulation of B-cell fate by antigen-receptor expression as novel prognostic indicators in chronic lymphocytic signals. Nat Rev Immunol 2002;2:945–56. leukemia. Blood 1999;94:1840–7. 6. Longo PG, Laurenti L, Gobessi S, Sica S, Leone G, Efremov DG. The 3. Hamblin TJ, Davis Z, Gardiner A, Oscier DG, Stevenson FK. Unmu- Akt/Mcl-1 pathway plays a prominent role in mediating antiapoptotic tated Ig V(H) genes are associated with a more aggressive form of signals downstream of the B-cell receptor in chronic lymphocytic chronic lymphocytic leukemia. Blood 1999;94:1848–54. leukemia B cells. Blood 2008;111:846–55.

www.aacrjournals.org Cancer Res; 70(18) September 15, 2010 7343

Hofbauer et al.

7. Petlickovski A, Laurenti L, Li XP, et al. Sustained signaling through 30. Wee KB, Surana U, Aguda BD. Oscillations of the p53-Akt network: the B-cell receptor induces Mcl-1 and promotes survival of chronic implications on cell survival and death. PLoS One 2009;4:e4407. lymphocytic leukemia B cells. Blood 2005;105:4820–7. 31. Eisele L, Haddad T, Sellmann L, Duhrsen U, Durig J. Expression 8. Barragan M, Iosillo BB, Campas C, Colomer D, Pons G, Gil J. levels of CD38 on leukemic B cells but not on non-leukemic T cells Involvement of protein kinase C and phosphatidylinositol 3-kinase are comparably stable over time and predict the course of disease in pathways in the survival of B-cell chronic lymphocytic leukemia cells. patients with chronic lymphocytic leukemia. Leuk Res 2009;33:775–8. Blood 2002;99:2969–76. 32. Gattei V, Bulian P, Del Principe MI, et al. Relevance of CD49d protein 9. Ringshausen I, Schneller F, Bogner C, et al. Constitutively activated expression as overall survival and progressive disease prognosticator phosphatidylinositol-3 kinase (PI-3K) is involved in the defect of in chronic lymphocytic leukemia. Blood 2008;111:865–73. apoptosis in B-CLL: association with protein kinase Cδ.Blood 33. Herling M, Patel KA, Weit N, et al. High TCL1 levels are a marker of 2002;100:3741–8. B-cell receptor pathway responsiveness and adverse outcome in 10. Zhuang J, Hawkins SF, Glenn MA, et al. Akt is activated in chronic- chronic lymphocytic leukemia. Blood 2009;114:4675–86. lymphocytic-leukemia cells and delivers a pro-survival signal: thera- 34. Pekarsky Y, Zanesi N, Aqeilan RI, Croce CM. Animal models for peutic potential of Akt inhibition. Haematologica 2010;95:110–8. chronic lymphocytic leukemia. J Cell Biochem 2007;100:1109–18. 11. de Frias M, Iglesias-Serret D, Cosialls AM, et al. Akt inhibitors induce 35. Tabrizi SJ, Niiro H, Masui M, et al. T cell leukemia/lymphoma 1 and apoptosis in chronic lymphocytic leukemia cells. Haematologica galectin-1 regulate survival/cell death pathways in human naive and 2009;94:1698–707. IgM+ memory B cells through altering balances in Bcl-2 family 12. Narducci MG, Virgilio L, Isobe M, et al. TCL1 oncogene activation in proteins. J Immunol 2009;182:1490–9. preleukemic T cells from a case of ataxia-telangiectasia. Blood 1995; 36. Stacchini A, Aragno M, Vallario A, et al. MEC1 and MEC2: two new 86:2358–64. cell lines derived from B-chronic lymphocytic leukaemia in prolym- 13. Said JW, Hoyer KK, French SW, et al. TCL1 oncogene expression in phocytoid transformation. Leuk Res 1999;23:127–36. B cell subsets from lymphoid hyperplasia and distinct classes of B 37. Hennessy BT, Smith DL, Ram PT, Lu Y, Mills GB. Exploiting the cell lymphoma. Lab Invest 2001;81:555–64. PI3K/AKT pathway for cancer drug discovery. Nat Rev Drug Discov 14. Herling M, Patel KA, Khalili J, et al. TCL1 shows a regulated expres- 2005;4:988–1004. sion pattern in chronic lymphocytic leukemia that correlates with 38. Cheng JQ, Lindsley CW, Cheng GZ, Yang H, Nicosia SV. The Akt/ molecular subtypes and proliferative state. Leukemia 2006;20:280–5. PKB pathway: molecular target for cancer drug discovery. Oncogene 15. Laine J, Kunstle G, Obata T, Sha M, Noguchi M. The protooncogene 2005;24:7482–92. TCL1 is an Akt kinase coactivator. Mol Cell 2000;6:395–407. 39. Barragan M, de Frias M, Iglesias-Serret D, et al. Regulation of Akt/ 16. Pekarsky Y, Koval A, Hallas C, et al. Tcl1 enhances Akt kinase activ- PKB by phosphatidylinositol 3-kinase-dependent and -independent ity and mediates its nuclear translocation. Proc Natl Acad Sci U S A pathways in B-cell chronic lymphocytic leukemia cells: role of protein 2000;97:3028–33. kinase Cβ. J Leukoc Biol 2006;80:1473–9. 17. Bichi R, Shinton SA, Martin ES, et al. Human chronic lymphocytic 40. Abrams ST, Lakum T, Lin K, et al. B-cell receptor signaling in chronic leukemia modeled in mouse by targeted TCL1 expression. Proc Natl lymphocytic leukemia cells is regulated by overexpressed active Acad Sci U S A 2002;99:6955–60. protein kinase CβII. Blood 2007;109:1193–201. 18. Hoyer KK, French SW, Turner DE, et al. Dysregulated TCL1 41. Kunstle G, Laine J, Pierron G, et al. Identification of Akt association promotes multiple classes of mature B cell lymphoma. Proc Natl and oligomerization domains of the Akt kinase coactivator TCL1. Mol Acad Sci U S A 2002;99:14392–7. Cell Biol 2002;22:1513–25. 19. Tazzari PL, Cappellini A, Bortul R, et al. Flow cytometric detection of total 42. Caligaris-Cappio F, Cignetti A, Granziero L, Ghia P. Chronic lympho- and serine 473 phosphorylated Akt. J Cell Biochem 2002;86:704–15. cytic leukaemia: a model for investigating potential new targets for 20. Sarbassov DD, Ali SM, Sengupta S, et al. Prolonged rapamycin treat- the therapy of indolent lymphomas. Best Pract Res Clin Haematol ment inhibits mTORC2 assembly and Akt/PKB. Mol Cell 2006;22: 2002;15:563–75. 159–68. 43. Lagneaux L, Delforge A, De Bruyn C, Bernier M, Bron D. Adhesion to 21. Thimmaiah KN, Easton JB, Germain GS, et al. Identification of N-10- bone marrow stroma inhibits apoptosis of chronic lymphocytic substituted phenoxazines as potent and specific inhibitors of Akt leukemia cells. Leuk Lymphoma 1999;35:445–53. signaling. J Biol Chem 2005;280:31924–35. 44. Kitada S, Andersen J, Akar S, et al. Expression of apoptosis-regulating 22. Ghia P, Chiorazzi N, Stamatopoulos K. Microenvironmental influ- proteins in chronic lymphocytic leukemia: correlations with in vitro and ences in chronic lymphocytic leukaemia: the role of antigen stimula- in vivo chemoresponses. Blood 1998;91:3379–89. tion. J Intern Med 2008;264:549–62. 45. Saxena A, Viswanathan S, Moshynska O, Tandon P, Sankaran K, 23. Holler C, Pinon JD, Denk U, et al. PKCβ is essential for the develop- Sheridan DP. Mcl-1 and Bcl-2/Bax ratio are associated with treatment of chronic lymphocytic leukemia in the TCL1 transgenic mouse ment response but not with Rai stage in B-cell chronic lymphocytic model: validation of PKCβ as a therapeutic target in chronic lympho- leukemia. Am J Hematol 2004;75:22–33. cytic leukemia. Blood 2009;113:2791–4. 46. Oscier DG, Gardiner AC, Mould SJ, et al. Multivariate analysis of 24. Kontny E, Kvanta A, Fredholm BB. Activation of protein kinase C and prognostic factors in CLL: clinical stage, IGVH gene mutational elevation of cAMP interact synergistically to raise c-Fos and AP-1 status, and loss or mutation of the p53 gene are independent prog- activity in Jurkat cells. Eur J Pharmacol 1992;227:333–8. nostic factors. Blood 2002;100:1177–84. 25. Alessi DR, Andjelkovic M, Caudwell B, et al. Mechanism of activation 47. Gryshchenko I, Hofbauer S, Stoecher M, et al. MDM2 SNP309 is of protein kinase B by insulin and IGF-1. EMBO J 1996;15:6541–51. associated with poor outcome in B-cell chronic lymphocytic leuke- 26. zum Buschenfelde CM, Wagner M, Lutzny G, et al. Recruitment of mia. J Clin Oncol 2008;26:2252–7. PKC-βII to lipid rafts mediates apoptosis-resistance in chronic lym- 48. Zenz T, Krober A, Scherer K, et al. Monoallelic TP53 inactivation is phocytic leukemia expressing ZAP-70. Leukemia 2010;24:141–52. associated with poor prognosis in chronic lymphocytic leukemia: 27. Wattel E, Preudhomme C, Hecquet B, et al. p53 mutations are asso- results from a detailed genetic characterization with long-term ciated with resistance to chemotherapy and short survival in hema- follow-up. Blood 2008;112:3322–9. tologic malignancies. Blood 1994;84:3148–57. 49. Uddin S, Hussain AR, Al Hussein KA, et al. Inhibition of phosphati- 28. Silber R, Degar B, Costin D, et al. Chemosensitivity of lymphocytes dylinositol 3′-kinase/AKT signaling promotes apoptosis of primary from patients with B-cell chronic lymphocytic leukemia to chlorambu- effusion lymphoma cells. Clin Cancer Res 2005;11:3102–8. cil, fludarabine, and camptothecin analogs. Blood 1994;84:3440–6. 50. Uriarte SM, Joshi-Barve S, Song Z, et al. Akt inhibition upregulates 29. Shankar E, Sivaprasad U, Basu A. Protein kinase Cε confers resis- FasL, downregulates c-FLIPs and induces caspase-8-dependent tance of MCF-7 cells to TRAIL by Akt-dependent activation of Hdm2 cell death in Jurkat T lymphocytes. Cell Death Differ 2005;12: and downregulation of p53. Oncogene 2008;27:3957–66. 233–42.

7344 Cancer Res; 70(18) September 15, 2010 Cancer Research

Modifying Akt Signaling in B-Cell Chronic Lymphocytic Leukemia Cells

Sebastian W. Hofbauer, Josefina D. Piñón, Gabriele Brachtl, et al.

Cancer Res Published OnlineFirst September 7, 2010.

Updated version Access the most recent version of this article at: doi:10.1158/0008-5472.CAN-09-4411

Supplementary Access the most recent supplemental material at: Material http://cancerres.aacrjournals.org/content/suppl/2010/09/03/0008-5472.CAN-09-4411.DC1

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/early/2010/09/03/0008-5472.CAN-09-4411. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.